The Phillips curve relates to the observed statistical relationship between inflation and unemployment.
In 1958, New Zealand economist AW Phillips published the results of his research into unemployment and inflation in the UK economy, from data gathered between 1861 and 1957.
Graphically, each ‘dot’ represents a year of data, with the Phillips curve the ‘line of best fit’ for the data.
When analysed, the data suggested a stable and inverse relationship between unemployment and inflation - at lower rates of unemployment, the inflation rate is higher.
Policy makers were quick to exploit the curve. If the economy was operating at point a, with unemployment troublingly high at 6%, the government would pump up demand with a fiscal stimulus. It could then predict that, sometime later, inflation would rise – in the graph, to 4%, as the economy moved to point b.
However, if attention switched to inflation, the government would reverse its policy and impose a fiscal constraint. The economy would then move back to point a. The process of periodically stimulating and constraining an economy was called ‘stop-go’ policy. This dominated policy in the UK from the 1950s to the 1980s.
The belief was that the policy worked through its effect on the labour market. Reflating by pumping in demand through a fiscal expansion (via a budget deficit) would cause the economy to expand temporarily, unemployment would fall creating a much 'tighter' labour market, with wages being driven up, followed by prices.
Conversely, a fiscal contraction with a 'surplus' budget created unemployment, with wages pegged back.
The Phillips curve provided a breakthrough in the understanding of the connection between inflation and unemployment - but by the 1970s, it had 'broken down' with seemingly no stable or inverse relationship. It was possible to have several inflation rates associated with a single unemployment rate.
For example, at a particular inflation rate of, say 6%, unemployment rates anywhere between 0 and 20% were all possible. No longer was it possible to set a given rate of inflation as a policy target and then engineer a specific expansion or contraction of an economy and accept a specific unemployment trade-off.
To resolve this theoretical problem, American economist Milton Friedman argued that there was a vertical long run Phillips curve, and a series of short-run curves - each associated with a different expectation of future inflation. So, what caused the breakdown of the Phillips curve?
As predicted, if, in the diagram below, the economy starts as point ‘A’, with 6% unemployment and 0% inflation, a short-run fiscal stimulus would move the economy to point ‘B’.
Unemployment would fall, at the cost of a higher rate of inflation. At some point a phenomenon called money illusion - where behaviour is irrational and responds to ‘nominal money values, rather than real values’ - would break down, and the economy would move to point ‘C’.
Unemployment rises back to 6%, but, as a consequence, higher inflation is now in the system.
Any further attempt to reflate the economy leads to accelerating inflation as the economy moves to point ‘D’, and then ‘E’ as money illusion breaks down.
The concept of money illusion dates back to American economist Irving Fisher, who defined money illusion as ‘failure to perceive that the dollar, or any other unit of money, expands or shrinks in value’ (Fisher,1928)1.
Given this, economic decisions may be seen as less rational in that changes in interest rates, inflation, money wages and welfare benefits may not lead to rational decision-making.
The breakdown of the Phillips curve relates to the breakdown of money illusion. Individuals who do not actively seek employment at a given real wage rate may then actively seek employment following an increase in nominal values - while real wages remain unchanged. Once inflation erodes the value of the increase in nominal wage, real wages return to their previous value.
However, once these workers perceive that this has happened, they will reverse their decision and no-longer actively look for work.
For example, an individual may remain inactive at a nominal wage of $200 per week. A rise in government expenditure to fund a local employment programme (funded largely by borrowing) may then see local wages rises to $220 during the year. This now provides the incentive for the individual to work.
However, with increased borrowing, prices start to rise by 10% in the year, and at the end of the year real wages fall back to $200. If money illusion then breaks down the individual realises he or she has been 'duped' and returns to being inactive!
Controversial as this idea seems, it shaped economic policy during the 1970s and 1980s as policy-makers looked to 're-educate' the public so that they would become more rational in how they reacted to changes in money wages.
The discovery of a vertical 'long run Phillips curve' at a specific level of unemployment cast doubt on the ability of policy-makers to achieve sustainable low levels of unemployment through the use of a fiscal stimulus, or indeed through expansionary monetary policy.
The consensus up until the late 1960s - derived from analysis using the Phillips curve - was that the level of unemployment could be driven down by the expansion of aggregate demand. This notion was questioned by several economists, including Milton Friedman and Edmund Phelps.
Friedman1 argued that, following expansionary policy, unemployment levels would always adjust back to their previous levels, at the 'natural rate of unemployment'.
Policy-makers tend to prefer to use the concept of the non-accelerating-inflation rate of unemployment ('NAIRU)' rather than 'natural rate'2. NAIRU is the rate of unemployment at which inflation stabilises. At an unemployment rate below this, inflation will accelerate and at a rate above NAIRU, inflation will decelerate.
It is now widely accepted that it is unlikely that fiscal policy on its own (or if at all) can permanently reduce unemployment below NAIRU.
If expansionary fiscal policy does - temporarily - reduce unemployment in the short run below its NAIRU rate it is likely to result in accelerating inflation, and is not sustainable in the long run. NAIRU can also be thought of a the 'sustainable rate of unemployment'.
For example, if NAIRU is at 6% (point a) a fiscal stimulus will simply result in a short-term reduction in unemployment to 4% (point b).
Once money illusion breaks down, the economy will move to point 'c' - with no long terms gain in reducing unemployment, but with a long term cost of higher inflation.
However, if the long run Phillips curve is shifted to the left, through supply-side policy, NAIRU can drop to 4% without inflationary pressure.
Supply-side policies, including reducing the level of welfare benefits and reducing marginal tax rates, would both reduce passive reliance on welfare benefits and make work more attractive, hence encouraging labour to join (or be retained in) the labour market without the need to raise wages (and increase costs, and inflationary pressure).
This recognition led economists and policy makers to focus on supply-side policy as an alternative to fiscal policy, and monetary policy, to reduce unemployment.
A cost shock, such as the oil price shock of the 1970s, and the more recent cost shock resulting from the Russia-Ukraine conflict, will shift the short run Phillips curve vertically upwards reflecting revised expectations regarding future inflation.
With inflation rates predicted to exceed 7% in many economies during 2022, real wages and living standards will fall which will encourage unions and non-unionised labour to push for increased wages.
As a result of the expectation of a fall in 'real' (inflation adjusted) profits, and following the increase in costs - including wage costs - firms are also likely to raise their prices. This can trigger a wage-price spiral, as occurred during the 1970s.
The solution is likely to involve increasingly tighter monetary policy to restrain demand and to revise inflationary expectations downwards.